CopyGov research guide
Did a congressional stock trade beat the S&P 500?
Compare disclosed stock purchases with SPY on matching dates, and distinguish stock gains, excess returns, and portfolio performance.
A positive return is only the first question
If a stock rises 10% while its benchmark rises 15% over the same dates, the stock gained money in the model but trailed by 5 percentage points. A headline showing only the 10% gain hides the relevant comparison. These numbers are illustrative.
CopyGov uses SPY as the market benchmark in its signal comparisons. SPY tracks the S&P 500; an ETF comparison is not identical to every possible version of an index-return series.
Match dates and read percentage points correctly
Compare the stock and benchmark from the same modeled entry through the same horizon. Subtract benchmark return from stock return to obtain excess return in percentage points. Do not describe this difference as a percentage increase in wealth.
The standard study horizons are 63, 126, and 252 trading sessions, labeled approximately three, six, and twelve months. Their actual end dates vary. A current-to-date result is a different measurement and should retain its own end date.
An average signal return is not a portfolio return
Signals can overlap, recur in the same stock, and come from different market periods. Averaging their returns does not specify how capital would have been allocated, how much cash was needed, or when positions were sold. It is not an annualized investment return.
Read the number of priced observations, missing observations, median excess return, and share beating SPY together. A high average can coexist with a negative median if a few observations dominate the result.
Test the selection rule
Choosing the best historical name after observing all of its returns is different from selecting a name before the next filing. A useful test chooses using only earlier information, then measures later results separately.
CopyGov’s existing walk-forward research does not establish dependable market outperformance. Use the evidence to investigate a claim, including when the evidence weakens it.